I obviously need to watch more top level football, because the angles involved in this particular argument are really, genuinely confusing me.
Let's hypothetically put the AAR is on the opposite side to the AR - aka, the opposite side to how it is currently done. If an incident occurs in the left hand side of the PA as viewed by the attacking team, where are you assuming the referee will be? I'm working on the assumption that the optimum position (with or without AARs) will be somewhere around the edge of the PA and somewhere left of centre. If this is the case you would often find that the AR is both unsighted and a long way away, but an AAR on this side would have a view almost perfectly 180 degrees different to that of the referee. That seems different enough that it adds value to the situation.
Unless the referee is going far far deeper into the PA that most I've ever noticed, he's going to have a very different view of any penalty/no penalty decision to the AAR on the goal line, looking from the opposite side. And even if referees do go a long way into the box, it's a fairly minor adjustment to their patrol path to tell them to end their runs a little shallower than normal and be more prepared in case of a counter as a result. I'm not talking about reinventing the wheel in terms of referee patrol paths.
And in the case of an incident on the right side of the PA, you've then got 3 officials looking at ~120 degree different angles - even if one of them has their view blocked, you're still adding more chance of a second viewpoint being clear.
As it is, if an incident occurs on the left side of the PA, the AAR in their current position could very easily be just as unsighted as the AR, leaving the decision entirely up the the referee. Meaning the AAR could often add nothing.